To expand a bit on the good info provided by Skeena88 and Matt B.
This question touches on one of the wonderful behaviors of our native salmonid; they typically exhibit complex and varied life histories to adapt to their environment and survival.
The ocean type juvenile has pretty specific threshold (about 3 inches) to achieve reasonable marine survival (currently about 1%). Typically, it takes 4 months or so of freshwater rearing to reach to achieve that threshold (for those that don't reach that threshold marine survival approaches zero). Within the ocean type there are two general rearing strategy as the fry emerge from the gravel, they begin seeking at rearing niches (a variety of habitats) for that freshwater rearing period. As the fry move downstream the move into the unoccupied niches essentially filling the preferred habitats. Historically and generally today there are more newly emerged fry than there are "rooms" in that preferred habitats. As those fry reach the estuaries (the interface between the freshwater and marine environments they have adopted to new rearing behaviors and preferred habitats to achieve the extra needed size to reach that survival threshold. While in that estuary habitat those fish grow pretty quickly (about a mm/day) depending on their size upon reaching that estuary zone typically requiring 4 to 6 weeks of extra reaching that needed size.
Of our native salmonids the Chinook are the only one where those small fry (less than 3 or 4 months of age) spend significant time (more than a week or two) are a critical strategy. The estuary effectively becomes the over-flow "rooms" potentially increasing the capacity of the basin to produce Chinook which of course why restoration of those loss estuary habitats is an important part of Chinook recovery planning.
For some stocks some of those fry that found in-river early rearing habitats remain in the river moving to more complex habitats and continuing to grow through that summer and winter before migrating the following spring as yearlings, having a stream life history. There is a marine survival advantage to having that larger size, but that growth comes at the cost of significant natural mortality (likely something in the 50% range). Having multiple rearing strategies and multiple ages of maturity provides protection for regular escapements and population stability.
How common is that "stream strategy? The ultimate test would be the portion of the NOR spawners had the stream rearing strategy. Don't know if we know that contribution was historically. Where we still find significant contributions of that stream strategy (Skagit and Snohomish) may provide some insight to that historical question. In the Skagit the spring stocks have surprisingly high contribution (Suiattle springs - 25 to 85%, upper Sauk 35 to 45%, Cascade 10 to 90%). For the summer stocks the upper Skagit stock has a 0 to 2% contribution of yearlings and the lower Sauk stock has 11 to 26%. One the Skagit there is one fall stock (lower Skagit) has 0 to 32% contribution. On there Snohomish the summer stock (Skykomish) has an average 20% contribution to the number of spawners and the Snoqualmie fall stock averages 16%.
As a side, in those two basins the fall stocks spawn significantly later than the typical Puget Sound fall stocks (Green river origin). The lower Skagit fish typically spawn through mid-November and the Snoqualmie fish spawn through November at times well into December (occasionally a few have been see spawning as late as Xmas.
Curt